Gross primary production estimation using remote sensing and meteorological variables

Abstract:

Accurate estimation of Gross Primary Production (GPP) is essential for understanding the climate change impact and global carbon cycle. Land surface remote sensing and meteorological variables have been used to estimate GPP. Satellite retrieved solar-induced chlorophyll fluorescence (SIF), as remote sensing variable, is considered a reliable indicator for GPP. However, SIF products has some limitations such as coarse e temporal or spatial resolutions, depending on which satellite product is used. Other remote sensing variables, such as the Normalized difference vegetation index (NDVI), as well as meteorological variables (such as temperature) are often used to quantify GPP. In this study, we evaluate the relationship between GPP and other variables based on the SIF product from Orbiting Carbon Observatory 2 (OCO-2) satellite, MODIS/Terra Gross Primary Productivity Gap-Filled (MOD17A2HGF) GPP product, MODIS/Terra Vegetation Indices Monthly (MOD13A3) product for NDVI and meteorological variables such as the maximum, minimum, and mean daily temperature, and focus on the corn belt as a case study. We analyze their relationships at both monthly and growing season scales and at the different spatial scales (0.1, 0.2 and 0.5 degree resolutions). The results show that monthly GPP data has the highest correlation with monthly SIF data. On the other hand, if we use growing season average data instead of monthly data, NDVI has the highest correlation with GPP. Although temperature in general has a very low correlation with GPP or SIF, the GPP/SIF relationship with temperature identifies a clear threshold pattern, reflecting the optimal growing temperature for crops and the threshold for the onset of heat stress, which is important to account for in developing a statistical model for GPP prediction.